Dynamic Range of Mass Accuracy in LTQ Orbitrap Hybrid Mass Spectrometer
Alexander Makarov, Eduard Denisov, Oliver Lange, and Stevan Horning Thermo Electron (Bremen) GmbH, Bremen, Germany
Using a novel orbitrap mass spectrometer, the authors investigate the dynamic range over which accurate masses can be determined (extent of mass accuracy) for short duration experiments typical for LC/MS. A linear ion trap is used to selectively fill an intermediate ion storage device (C-trap) with ions of interest, following which the ensemble of ions is injected into an orbitrap mass analyzer and analyzed using image current detection and fast Fourier transformation. Using this technique, it is possible to generate ion populations with intraspec- trum intensity ranges up to 104. All measurements (including ion accumulation and image current detection) were performed in less than1sataresolving power of 30,000. It was shown that 5-ppm mass accuracy of the orbitrap mass analyzer is reached with Ͼ95% probability at a dynamic range of more than 5000, which is at least an order of magnitude higher than typical values for time-of-flight instruments. Due to the high resolving power of the orbitrap, accurate mass of an ion could be determined when the signal was reliably distinguished from noise Ͼ ѧ (S/Np-p 2 3). (J Am Soc Mass Spectrom 2006, 17, 977–982) © 2006 American Society for Mass Spectrometry
he dynamic range over which accurate measure- troiding introduced by the noise of the image current ments of mass can be made (“extent of mass preamplifier [5– 8]. Unlike TOFs, FT ICR employs much Taccuracy”) is a key analytical figure-of-merit for slower acquisition systems with much higher dynamic any accurate-mass analyzer. In practice, such analyzers range. At high intensities, Coulomb repulsion, rather are coupled to liquid chromatography or other separa- than detector saturation, produces mass shifts that tion methods, and measurements are made for transient depend not only on the total charge [9 –11] but also on signals (e.g., with spectra recorded at a rate of 1 the intensities of individual mass peaks [12, 13]. Over- spectrum/s). For any analyzer, mass accuracy is limited all, intrascan dynamic range of a few thousand is statistically by too few ions detected or by peak position possible [8, 14] with mass accuracy of a few ppm. shifts due to too many ions. Though being a universal This work investigates the extent of mass accuracy problem, limitations to the extent of mass accuracy have for a novel Fourier transform mass spectrometer: been investigated in detail for time-of-flight (TOF) mass LTQ Orbitrap. This instrument combines a linear ion analyzers. These analyzers are particularly susceptible trap with radial ejection [15] and an orbitrap mass to variations in ion intensity because they use fast analyzer [16]. The orbitrap mass analyzer is an elec- acquisition systems with inherently modest dynamic trostatic trap wherein tangentially injected ions rotate range [1– 4]. Even with lock-mass and analyte signals around a central electrode, being confined by apply- far from saturation, a strong dependence of peak posi- ing an appropriate voltage between the outer and tion on peak intensity is observed. As a result, 5 ppm central electrodes. Mass analysis is based on image r.m.s. mass accuracy cannot be achieved over a signal current detection of frequencies of axial oscillations. range larger than a few hundred in 1 s acquisition in ion Therefore, its extent of mass accuracy is limited by counting TOFs, even when advanced algorithms for the same factors as FT ICR. The objective of this work intensity correction are employed [1–3]. TOFs with is to determine upper and lower limits of ion inten- analog detection are in principle capable of a dynamic sities for accurate mass analysis and ways for their range of around one thousand [4]. improvement. In Fourier transform ion cyclotron resonance (FT ICR) mass spectrometry, mass accuracy at low signal intensities is limited by the imprecision of peak cen- Experimental All experiments were carried out using a mixture of Published online June 5, 2006 Ultramark 1600 (Lancaster Synthesis Inc., Windham, Address reprint requests to Dr. A. A. Makarov, Thermo Electron (Bremen) GmbH, Hanna Kunath Strasse 11, Bremen 28199, Germany. E-mail: NH) and MRFA peptide in 50:50 vol/vol water/aceto- [email protected] nitrile solution.
© 2006 American Society for Mass Spectrometry. Published by Elsevier Inc. Received January 13, 2006 1044-0305/06/$32.00 Revised March 16, 2006 doi:10.1016/j.jasms.2006.03.006 Accepted March 16, 2006 978 MAKAROV ET AL. J Am Soc Mass Spectrom 2006, 17, 977–982
Figure 1. Experimental sequence for measurements of the extent of mass accuracy the orbitrap mass analyzer: (a) Injection of the first set of ions and trapping in the C-trap; (b) injection of the second set of ions and trapping in the C-trap; (c) pulsed injection of mixed ion population into the orbitrap; (d) ion detection in the orbitrap.
Results and Discussion and an orbitrap mass analyzer. Key to operation of this system is a C-shaped storage trap, which is used to Instrument Operation store and collisionally cool ions before injection into the The mass spectrometer depicted in Figure 1 is a hybrid orbitrap. With this device, ions are pulsed into the system combining a linear ion trap mass spectrometer central point of the C-trap arc that coincides with the J Am Soc Mass Spectrom 2006, 17, 977–982 DYNAMIC RANGE OF LTQ ORBITRAP 979
DR_524low_1e6L_int_profile # 24 RT: 0.39 AV: 1 NL: 1.49E10 T: FTMS + p ESI Full ms2 [email protected] [ 400.00-2000.00] 1421.98035 R=19245 100
90 0.06
0.05 80 0.04 Relative Abundance Relative
70 0.03 524.26611 R=30209 60 0.02
0.01 50 0.00 521 522 523 524 525 526 527 40 m/z Relative Abundance Relative
30
20
10
0 400 600 800 1000 1200 1400 1600 1800 2000 m/z
Figure 2. A typical mass spectrum acquired at the ratio 5000:1 of maximum to minimum intensities (external calibration). orbitrap entrance aperture. Ions are captured in the ion populations including, if necessary, internal cali- orbitrap by rapidly increasing the electric field and brants. Mass calibration coefficients were determined detection of image current from coherent ion packets for different AGC target values and interpolated for takes place after voltages have stabilized [16]. Signals intermediate values. No intensity-dependant correc- from each of the orbitrap outer electrodes are amplified tions of m/z were made for data processing. by a differential amplifier and transformed into a fre- The resolving power was reduced to nominal 30,000 quency spectrum by fast Fourier transformation. The (at m/z 400 Th after zero-filling and Kaiser-Bessel apo- frequency spectrum is converted into a mass spectrum dization, 0.38 s transient duration) so that the experi- using a two-point calibration and processed with Xcali- ment cycle time of 1 s still allowed more than sufficient bur software. time to store up to a million of ions in the C-trap. All data below correspond to a single spectrum acquisition. Measurement Methodology To model the widest possible range of conditions, intensities of dominant and minor peaks were varied To explore the extent of mass accuracy of the orbitrap over orders of magnitude to achieve variations of ratio analyzer, it is important to provide a reproducible and of intensities between 1 and 10,000. For internal calibra- as wide as possible spread of signal intensities within tion evaluation, the intense peak was used as the the same spectrum. An effort was made to achieve a calibrant. Here and below, all resolving powers are wide range of signal intensities by using an electrospray presented as full-width half-maximum (FWHM) values. source with widely different concentrations of analytes. However, it appeared that competition between ana- Results of Measurements lytes for protons precluded the electrospray source from producing reliable and controllable signals at Figure 2 shows a typical mass spectrum used to deter- levels in the range 1:2000 to 1:5000, compared with the mine mass errors at the extreme limits of dynamic major component of the analyte mixture. To investigate range. Target values were 106 for the major peak and properties of the orbitrap analyzer rather than the 102 for the minor peak; however, space charge repulsion electrospray source, another approach was applied. in the C-trap resulted in a significant reduction of the This approach capitalized on the ability of the C-trap to major peak intensity (about 2-fold). The minor peak has store multiple fills from the linear ion trap per injection such a low S/N that noise starts to limit the precision of into the orbitrap, as illustrated in Figure 1. The number mass measurement in agreement with the published of ions in each fill is individually controlled over several literature [5–7]. It this paper, noise is characterized as orders of magnitude using automatic gain control the maximum peak-to-peak amplitude of thermal noise (AGC), while the selection of masses in each fill is of the preamplifier over the full mass range (for exam- regulated using isolation in the linear ion trap. This ple, in Figure 2 the noise stays at 0.008% of the major creates a flexible and versatile tool for forming desired peak). For a signal with a low S/N ratio, it was found 980 MAKAROV ET AL. J Am Soc Mass Spectrom 2006, 17, 977–982
External calibration Internal calibration S/N≈2 S/N≈2 DR=5,000 DR=5,000
10.0 10.0 a) 9.0 b) 9.0 8.0 m/z 524.2650 8.0 m/z 524.2650 7.0 7.0
6.0 6.0
5.0 5.0 error, ppm 4.0 4.0 Mass error, ppm Mass 3.0 3.0
2.0 2.0
1.0 1.0
0.0 0.0 1 10 100 1000 10000 1 10 100 1000 10000 Intensity ratio Intensity ratio c) 10.0 10.0 9.0 d) 9.0 8.0 m/z 1121.9970 8.0 m/z 1121.9970 7.0 7.0 m 6.0 6.0
5.0 5.0 rror, ppm error, pp 4.0 4.0 Mass e 3.0 Mass 3.0
2.0 2.0
1.0 1.0
0.0 0.0 1 10 100 1000 10000 1 10 100 1000 10000 IntensityMax/min intensity ratio Max/minIntensity intensity ratio
10.0 10.0 e) 9.0 f) 9.0 8.0 m/z 1721.9587 8.0 m/z 1721.9587 7.0 7.0
6.0 6.0 ppm
5.0 ror, 5.0 er 4.0 4.0 ss error, ppm Ma 3.0 Mass 3.0
2.0 2.0
1.0 1.0
0.0 0.0 1 10 100 1000 10000 1 10 100 1000 10000 Intensity ratio ratio
Figure 3. Absolute measured mass errors (points) and trend of r.m.s. mass error versus ratio of intensities between the major peak at m/z 1421.97,786 (lines) and peaks at different m/z:(a) m/z ϭ 524.2650 at resolving power R ϭ 30,000, external calibration; (b) m/z ϭ 524.2650 at resolving power R ϭ 30,000, internal calibration; (c) m/z ϭ 1121.9970 at resolving power R ϭ 21,000, external calibration; (d) m/z ϭ 1121.9970 at resolving power R ϭ 21,000, internal calibration; (e) m/z ϭ 1721.9587 at resolving power R ϭ 17,400, external calibration; (f) m/z ϭ 1721.9587 at resolving power R ϭ 17,400, internal calibration. that lower resolving power results in mass errors in- minor and major peaks. In all these measurements, creasing almost proportionally. Moreover, both signal the major peak at nominal mass 1422 was dosed with and noise increase with acquisition time, but signal a smaller analyte with m/z indicated in the figure increases in proportion to time, while noise increases in along with the ratio of their signal intensities on the proportion to the square root of time. Thus, S/N abscissa axis. The major peak was also used as a improves with acquisition time. For this reason, it is calibrant to produce data for internal calibration desirable to use longer acquisitions. (Figure 3b, d, f). At lower values of maximum/ Figure 3 brings together numerous results of mea- minimum intensity (below 100), target numbers for surements for different m/z and target values for both both major and minor peaks were varied so that the J Am Soc Mass Spectrom 2006, 17, 977–982 DYNAMIC RANGE OF LTQ ORBITRAP 981
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5 m 3.0
2.5
2.0
1.5 on, pp 1.0
ati 0.5
vi 0.0
-0.5 De -1.0
-1.5
-2.0
-2.5
-3.0
-3.5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Time, hours
Figure 4. Long-term stability of accurate mass measurements with a large difference in signal intensities and external mass calibration (green trace: m/z 1421.97,786 at 100%; blue trace: m/z 524.26,496 at Ͻ0.02%). total target number of stored ions was varied be- 4, as shown by two mass traces with an intensity ratio tween 5·104 and 106. of ϳ5000, which remained stable for more than 20 h. It was observed that for resolving powers of 30,000 In this plot, one point was acquired every 6 s. It and higher (Figure 3a, b), mass errors for all 1100 data should be noted that mass measurement variations points are under 5 ppm. For resolving powers around for the minor component is much higher because of and below 20,000 (which are observed for higher m/z at low S/N. Spikes on the blue trace probably result the same resolution setting due to a lower number of from intermittent variations of minor component oscillations over the acquisition time), several measure- intensity, i.e., lower than usual S/N ratios due to ments show mass errors in the range 4 to 7 ppm (Figure statistical effects. The overall trend demonstrates 3c–f). However, for all these resolving powers, root- stability and effectiveness of thermal regulation, mean-square mass errors stay well within 5 ppm down which is required for external calibration. ϭ to S/Np-p 2. With the major component acquired at a 6 ϭ target number 10 , the S/Np-p 2 threshold is well Conclusions below 0.02% (i.e., 1:5000 intensity ratio); thus the extent of mass accuracy of the orbitrap analyzer extends at The extent of mass accuracy determines the true utility least up to 5000. of the accurate mass capability of a mass spectrometer At low S/N (i.e., large max/min intensity ratios), for real-life applications, much more than other param- noise is the main contributor to the error of peak eters (even resolving power). From this point of view, centroiding, so there is no significant difference in mass the LTQ Orbitrap enables accurate mass measurements errors between internal and external calibrations. At over an intensity range of 5000 that matches or exceeds higher S/N ratios (small max/min intensity ratios), the range of signal intensities in the electrospray ion precision of mass measurements improves, and for source [18] when operated with liquid separations. internal calibration it remains limited only by the accu- Further improvements of extent of mass accuracy racy of the two-point mass calibration. Graphs on can be made at both extremes of signal intensities. For Figure 3 b, d, and f show that internal mass accuracy low intensities, lower capacitance of the orbitrap and can be better than 1 ppm, which agrees well with lower thermal noise of the image current preamplifier published measurements for peptide mixtures [17]. are needed to reduce noise and, thus, allow lower For external mass accuracy, small variations of the intensity signals to be detected. At high intensities, the output of the high-voltage power supply over time number of injected ions appeared to saturate at levels cause an increase of mass errors, without affecting that are still well within the space charge capacity of the mass errors for internal mass accuracy. This is espe- orbitrap. Thus, further improvement of the C-trap cially pronounced for longer durations of measure- would be required to approach the inherent space ments like in Figure 3a, where several acquisitions charge limit of the orbitrap, i.e., space charge when were separated by Ͼ1 h from each other. The stability mass accuracy starts to drop noticeably due to Coulomb of external mass calibration is demonstrated in Figure repulsion. 982 MAKAROV ET AL. J Am Soc Mass Spectrom 2006, 17, 977–982
8. Limbach, P. A.; Grosshans, P. B.; Marshall, A. G. Experimental deter- Acknowledgments mination of the number of trapped ions, detection limit, and dynamic range in Fourier-transform ion-cyclotron resonance mass spectrometry. The authors express their deep gratitude to colleagues from the Anal. Chem. 1993, 65, 135–140. development team of LTQ Orbitrap instrument: Wilko Balschun, 9. Easterling, M. L.; Mize, T. H.; Amster, I. J. Routine part-per-million mass Alexander Kholomeev, Dr. Kerstin Strupat, Dr. Reinhold Pesch, accuracy for high-mass ions: Space-charge effects in MALDI FT-ICR. Anal. Chem. 1999, 71, 624–632. Frank Czemper, Oliver Hengelbrock of Thermo Electron, Bremen, 10. Ledford, E. B.; Rempel, D. L.; Gross, M. L. Space-charge effects in Germany, and Dr. Eric Hemenway of Thermo Electron, San Jose, Fourier-transform mass spectrometry-mass calibration. Anal. Chem. California, USA. The authors are also grateful to Dr. Michael 1984, 56, 2744–2748. Senko for his valuable comments. 11. Jeffries, J. B.; Barlow, S. E.; Dunn, G. H. Theory of space-charge shift of ion-cyclotron resonance frequencies. Int. J. Mass Spectrom. Ion Processes 1983, 54, 169–187. 12. Brown, C. E.; Smith, M. J. The present status and prospects for Fourier References transform-ion cyclotron resonance/mass spectrometry. Spectrosc. World 1990, 1, 25–30. 1. Blom, K. F. Estimating the precision of exact mass measurements on an 13. Masselon, C.; Tolmachev, A. V.; Anderson, G. A.; Harkewicz, R.; Smith, orthogonal time-of-flight mass spectrometer. Anal. Chem. 2001, 73, R. D. Mass measurement errors caused by “local” frequency perturba- 715–719. tions in FTICR mass spectrometry. J. Am. Soc. Mass Spectrom. 2002, 13, 2. Wu, J.; McAllister H. Exact mass measurement on an electrospray 99–106. ionization time-of-flight mass spectrometer: Error distribution and 14. Syka, J. E. P.; Marto, J. A.; Bai, D. L.; Horning, S.; Senko, M. W.; selective averaging. J. Mass Spectrom. 2003, 38, 1043–1053. Schwartz, J. C.; Ueberheide, B.; Garcia, B.; Busby, S.; Muratore, T.; 3. Colombo, M.; Sirtori, F. R.; Rizzo, V. A fully automated method for Shabanowitz, J.; Hunt, D. F. Novel linear quadrupole ion trap/FT mass accurate mass determination using high-performance liquid chroma- spectrometer: Performance characterization and use in the comparative tography with a quadrupole/orthogonal acceleration time-of-flight analysis of histone H3 post-translational modifications. J. Proteome Res. mass spectrometer. Rapid Commun. Mass Spectrom. 2004, 18, 511–517. 2004, 3, 621–626. 4. Fjeldsted, J. Time-of-Flight Mass Spectrometry, Technical Overview; Report 15. Schwartz, J. C.; Senko, M. W.; Syka, J. E. P. A two-dimensional 5989-0373EN, Agilent Technologies: 2003; (http://www.chem.agilent. quadrupole ion trap mass spectrometer. J. Am. Soc. Mass Spectrom. 2002, com) 13, 659–669. 5. Chen, L.; Cottrell, C. E.; Marshall, A. G. Effect of signal-to-noise ratio 16. Makarov, A. A. Electrostatic axially-harmonic orbital trapping: A novel and number of data points upon precision in measurement of peak high-performance technique of mass analysis. Anal. Chem. 2000, 72, amplitude, position, and width in Fourier transform spectrometry. 1156–1162. Chemom. Intell. Lab. Syst. 1986, 1, 51–58. 17. Olsen, J. V.; de Godoy, L. M.; Li, G.; Macek, B.; Mortensen, P.; Pesch, R.; 6. Verdun, F. R.; Giancaspro, C.; Marshall, A. G. Effects of noise, time- Makarov, A. A.; Lange, O.; Horning, S.; Mann, M. Parts per million domain damping, zero-filling, and the FFT algorithm on the “exact” mass accuracy on an orbitrap mass spectrometer via lock-mass injection interpolation of fast Fourier transform spectra. Appl. Spectrosc. 1988, 42, into a C-trap. Mol. Cell. Proteom. 2005, 4, 2010–2021. 715–721. 18. Tang, K.; Page, J. S.; Smith, R. D. Charge competition and the linear 7. Marshall, A. G.; Verdun, F. R. Fourier Transforms in NMR, Optical, and dynamic range of detection in electrospray ionization mass spectrom- Mass Spectrometry; Elsevier: Amsterdam, 1990; pp 150–155. etry. J. Am. Soc. Mass Spectrom. 2004, 15, 1416–1423.